A Hybrid Newton–Raphson and Particle Swarm Optimization Method for Target Motion Analysis by Batch Processing

Abstract

Bearing-only target motion analysis (BO-TMA) by batch processing remains a challenge due to the lack of information on underwater target maneuvering and the nonlinearity of sensor measurements. Traditional batch estimation for BO-TMA is mainly performed based on deterministic algorithms, and studies performed with heuristic algorithms have recently been reported. However, since the two algorithms have their own advantages and disadvantages, interest in a hybrid method that complements the disadvantages and combines the advantages of the two algorithms is increasing. In this study, we proposed Newton–Raphson particle swarm optimization (NRPSO): a hybrid method that combines the Newton–Raphson method and the particle swarm optimization method, which are representative methods that utilize deterministic and heuristic algorithms, respectively. The BO-TMA performance obtained using the proposed NRPSO was tested by varying the measurement noise and number of measurements for three targets with different maneuvers. The results showed that the advantages of both methods were well combined, which improved the performance.

Underwater target passive acoustic localization method based on Hanbury Brown–Twiss interference

“…TMA is a remote passive localization method based on a plane-wave propagation model. In 2021, Oh et al (2021) proposed Newton–Raphson particle swarm optimization based on the advantages of complementary deterministic and heuristic algorithms, which improved the performance of bearing-only TMA. However, TMA requires dynamic information about the relative motion between the positioning system and the target, which requires a relatively large amount of calculation (Kim, 2016; Kim et al , 2017; Nguyen and Dogancay, 2017).…”

Intelligence-Aware Batch Processing for TMA with Bearings-Only Measurements

“…Other relevant approaches in the literature include, among other works: applications to sensor network localization [19]; algorithms based on direction-of-arrival measurements, modeled by von Mises-Fisher distributions [20]; pseudolinear estimators for 3D target motion analysis by a single moving ownship collecting azimuth and elevation angle measurements [21]; a methodology based on a bank of batch maximum a posteriori (MAP) estimators as a general estimation framework that provides the relinearization of the entire state trajectory, multihypothesis tracking and an efficient hypothesis generation scheme [22]; an approach based on Newton-Raphson methods and Particle Swarm Optimization [23]; a methodology to combine target motion compensation and track-before-detect methods within passive radar based on global navigation satellite systems (GNSS) for the detection of maritime targets [24]; TMA from cosines of conical angles acquired by a towed array [25]; and a new pseudolinear filter for bearings-only tracking without the requirement of bias compensation [26].…”